1 /*-
2 * SPDX-License-Identifier: BSD-3-Clause
3 *
4 * Copyright (C) 1995, 1996, 1997, and 1998 WIDE Project.
5 * All rights reserved.
6 *
7 * Redistribution and use in source and binary forms, with or without
8 * modification, are permitted provided that the following conditions
9 * are met:
10 * 1. Redistributions of source code must retain the above copyright
11 * notice, this list of conditions and the following disclaimer.
12 * 2. Redistributions in binary form must reproduce the above copyright
13 * notice, this list of conditions and the following disclaimer in the
14 * documentation and/or other materials provided with the distribution.
15 * 3. Neither the name of the project nor the names of its contributors
16 * may be used to endorse or promote products derived from this software
17 * without specific prior written permission.
18 *
19 * THIS SOFTWARE IS PROVIDED BY THE PROJECT AND CONTRIBUTORS ``AS IS'' AND
20 * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
21 * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
22 * ARE DISCLAIMED. IN NO EVENT SHALL THE PROJECT OR CONTRIBUTORS BE LIABLE
23 * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
24 * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
25 * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
26 * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
27 * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
28 * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
29 * SUCH DAMAGE.
30 */
31
32 /*-
33 * Copyright (c) 1982, 1986, 1988, 1993
34 * The Regents of the University of California.
35 * All rights reserved.
36 *
37 * Redistribution and use in source and binary forms, with or without
38 * modification, are permitted provided that the following conditions
39 * are met:
40 * 1. Redistributions of source code must retain the above copyright
41 * notice, this list of conditions and the following disclaimer.
42 * 2. Redistributions in binary form must reproduce the above copyright
43 * notice, this list of conditions and the following disclaimer in the
44 * documentation and/or other materials provided with the distribution.
45 * 3. Neither the name of the University nor the names of its contributors
46 * may be used to endorse or promote products derived from this software
47 * without specific prior written permission.
48 *
49 * THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND
50 * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
51 * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
52 * ARE DISCLAIMED. IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE
53 * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
54 * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
55 * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
56 * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
57 * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
58 * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
59 * SUCH DAMAGE.
60 *
61 * @(#)raw_ip.c 8.2 (Berkeley) 1/4/94
62 */
63
64 #include <sys/cdefs.h>
65 __FBSDID("$FreeBSD: stable/12/sys/netinet6/raw_ip6.c 370289 2021-08-11 02:35:48Z kbowling $");
66
67 #include "opt_ipsec.h"
68 #include "opt_inet6.h"
69
70 #include <sys/param.h>
71 #include <sys/errno.h>
72 #include <sys/jail.h>
73 #include <sys/kernel.h>
74 #include <sys/lock.h>
75 #include <sys/malloc.h>
76 #include <sys/mbuf.h>
77 #include <sys/priv.h>
78 #include <sys/proc.h>
79 #include <sys/protosw.h>
80 #include <sys/signalvar.h>
81 #include <sys/socket.h>
82 #include <sys/socketvar.h>
83 #include <sys/sx.h>
84 #include <sys/syslog.h>
85
86 #include <net/if.h>
87 #include <net/if_var.h>
88 #include <net/if_types.h>
89 #include <net/route.h>
90 #include <net/vnet.h>
91
92 #include <netinet/in.h>
93 #include <netinet/in_var.h>
94 #include <netinet/in_systm.h>
95 #include <netinet/in_pcb.h>
96
97 #include <netinet/icmp6.h>
98 #include <netinet/ip6.h>
99 #include <netinet/ip_var.h>
100 #include <netinet6/ip6protosw.h>
101 #include <netinet6/ip6_mroute.h>
102 #include <netinet6/in6_pcb.h>
103 #include <netinet6/ip6_var.h>
104 #include <netinet6/nd6.h>
105 #include <netinet6/raw_ip6.h>
106 #include <netinet6/scope6_var.h>
107 #include <netinet6/send.h>
108
109 #include <netipsec/ipsec_support.h>
110
111 #include <machine/stdarg.h>
112
113 #define satosin6(sa) ((struct sockaddr_in6 *)(sa))
114 #define ifatoia6(ifa) ((struct in6_ifaddr *)(ifa))
115
116 /*
117 * Raw interface to IP6 protocol.
118 */
119
120 VNET_DECLARE(struct inpcbhead, ripcb);
121 VNET_DECLARE(struct inpcbinfo, ripcbinfo);
122 #define V_ripcb VNET(ripcb)
123 #define V_ripcbinfo VNET(ripcbinfo)
124
125 extern u_long rip_sendspace;
126 extern u_long rip_recvspace;
127
128 VNET_PCPUSTAT_DEFINE(struct rip6stat, rip6stat);
129 VNET_PCPUSTAT_SYSINIT(rip6stat);
130
131 #ifdef VIMAGE
132 VNET_PCPUSTAT_SYSUNINIT(rip6stat);
133 #endif /* VIMAGE */
134
135 /*
136 * Hooks for multicast routing. They all default to NULL, so leave them not
137 * initialized and rely on BSS being set to 0.
138 */
139
140 /*
141 * The socket used to communicate with the multicast routing daemon.
142 */
143 VNET_DEFINE(struct socket *, ip6_mrouter);
144
145 /*
146 * The various mrouter functions.
147 */
148 int (*ip6_mrouter_set)(struct socket *, struct sockopt *);
149 int (*ip6_mrouter_get)(struct socket *, struct sockopt *);
150 int (*ip6_mrouter_done)(void);
151 int (*ip6_mforward)(struct ip6_hdr *, struct ifnet *, struct mbuf *);
152 int (*mrt6_ioctl)(u_long, caddr_t);
153
154 /*
155 * Setup generic address and protocol structures for raw_input routine, then
156 * pass them along with mbuf chain.
157 */
158 int
rip6_input(struct mbuf ** mp,int * offp,int proto)159 rip6_input(struct mbuf **mp, int *offp, int proto)
160 {
161 struct ifnet *ifp;
162 struct mbuf *m = *mp;
163 struct ip6_hdr *ip6 = mtod(m, struct ip6_hdr *);
164 struct inpcb *inp;
165 struct inpcb *last = NULL;
166 struct mbuf *opts = NULL;
167 struct sockaddr_in6 fromsa;
168 struct epoch_tracker et;
169
170 RIP6STAT_INC(rip6s_ipackets);
171
172 init_sin6(&fromsa, m, 0); /* general init */
173
174 ifp = m->m_pkthdr.rcvif;
175
176 INP_INFO_RLOCK_ET(&V_ripcbinfo, et);
177 CK_LIST_FOREACH(inp, &V_ripcb, inp_list) {
178 /* XXX inp locking */
179 if ((inp->inp_vflag & INP_IPV6) == 0)
180 continue;
181 if (inp->inp_ip_p &&
182 inp->inp_ip_p != proto)
183 continue;
184 if (!IN6_IS_ADDR_UNSPECIFIED(&inp->in6p_laddr) &&
185 !IN6_ARE_ADDR_EQUAL(&inp->in6p_laddr, &ip6->ip6_dst))
186 continue;
187 if (!IN6_IS_ADDR_UNSPECIFIED(&inp->in6p_faddr) &&
188 !IN6_ARE_ADDR_EQUAL(&inp->in6p_faddr, &ip6->ip6_src))
189 continue;
190 if (last != NULL) {
191 struct mbuf *n = m_copym(m, 0, M_COPYALL, M_NOWAIT);
192
193 #if defined(IPSEC) || defined(IPSEC_SUPPORT)
194 /*
195 * Check AH/ESP integrity.
196 */
197 if (IPSEC_ENABLED(ipv6)) {
198 if (n != NULL &&
199 IPSEC_CHECK_POLICY(ipv6, n, last) != 0) {
200 m_freem(n);
201 /* Do not inject data into pcb. */
202 n = NULL;
203 }
204 }
205 #endif /* IPSEC */
206 if (n) {
207 if (last->inp_flags & INP_CONTROLOPTS ||
208 last->inp_socket->so_options & SO_TIMESTAMP)
209 ip6_savecontrol(last, n, &opts);
210 /* strip intermediate headers */
211 m_adj(n, *offp);
212 if (sbappendaddr(&last->inp_socket->so_rcv,
213 (struct sockaddr *)&fromsa,
214 n, opts) == 0) {
215 soroverflow(last->inp_socket);
216 m_freem(n);
217 if (opts)
218 m_freem(opts);
219 RIP6STAT_INC(rip6s_fullsock);
220 } else
221 sorwakeup(last->inp_socket);
222 opts = NULL;
223 }
224 INP_RUNLOCK(last);
225 last = NULL;
226 }
227 INP_RLOCK(inp);
228 if (__predict_false(inp->inp_flags2 & INP_FREED))
229 goto skip_2;
230 if (jailed_without_vnet(inp->inp_cred)) {
231 /*
232 * Allow raw socket in jail to receive multicast;
233 * assume process had PRIV_NETINET_RAW at attach,
234 * and fall through into normal filter path if so.
235 */
236 if (!IN6_IS_ADDR_MULTICAST(&ip6->ip6_dst) &&
237 prison_check_ip6(inp->inp_cred,
238 &ip6->ip6_dst) != 0)
239 goto skip_2;
240 }
241 if (inp->in6p_cksum != -1) {
242 RIP6STAT_INC(rip6s_isum);
243 if (m->m_pkthdr.len - (*offp + inp->in6p_cksum) < 2 ||
244 in6_cksum(m, proto, *offp,
245 m->m_pkthdr.len - *offp)) {
246 RIP6STAT_INC(rip6s_badsum);
247 /*
248 * Drop the received message, don't send an
249 * ICMP6 message. Set proto to IPPROTO_NONE
250 * to achieve that.
251 */
252 proto = IPPROTO_NONE;
253 goto skip_2;
254 }
255 }
256 /*
257 * If this raw socket has multicast state, and we
258 * have received a multicast, check if this socket
259 * should receive it, as multicast filtering is now
260 * the responsibility of the transport layer.
261 */
262 if (inp->in6p_moptions &&
263 IN6_IS_ADDR_MULTICAST(&ip6->ip6_dst)) {
264 /*
265 * If the incoming datagram is for MLD, allow it
266 * through unconditionally to the raw socket.
267 *
268 * Use the M_RTALERT_MLD flag to check for MLD
269 * traffic without having to inspect the mbuf chain
270 * more deeply, as all MLDv1/v2 host messages MUST
271 * contain the Router Alert option.
272 *
273 * In the case of MLDv1, we may not have explicitly
274 * joined the group, and may have set IFF_ALLMULTI
275 * on the interface. im6o_mc_filter() may discard
276 * control traffic we actually need to see.
277 *
278 * Userland multicast routing daemons should continue
279 * filter the control traffic appropriately.
280 */
281 int blocked;
282
283 blocked = MCAST_PASS;
284 if ((m->m_flags & M_RTALERT_MLD) == 0) {
285 struct sockaddr_in6 mcaddr;
286
287 bzero(&mcaddr, sizeof(struct sockaddr_in6));
288 mcaddr.sin6_len = sizeof(struct sockaddr_in6);
289 mcaddr.sin6_family = AF_INET6;
290 mcaddr.sin6_addr = ip6->ip6_dst;
291
292 blocked = im6o_mc_filter(inp->in6p_moptions,
293 ifp,
294 (struct sockaddr *)&mcaddr,
295 (struct sockaddr *)&fromsa);
296 }
297 if (blocked != MCAST_PASS) {
298 IP6STAT_INC(ip6s_notmember);
299 goto skip_2;
300 }
301 }
302 last = inp;
303 continue;
304 skip_2:
305 INP_RUNLOCK(inp);
306 }
307 INP_INFO_RUNLOCK_ET(&V_ripcbinfo, et);
308 #if defined(IPSEC) || defined(IPSEC_SUPPORT)
309 /*
310 * Check AH/ESP integrity.
311 */
312 if (IPSEC_ENABLED(ipv6) && last != NULL &&
313 IPSEC_CHECK_POLICY(ipv6, m, last) != 0) {
314 m_freem(m);
315 IP6STAT_DEC(ip6s_delivered);
316 /* Do not inject data into pcb. */
317 INP_RUNLOCK(last);
318 } else
319 #endif /* IPSEC */
320 if (last != NULL) {
321 if (last->inp_flags & INP_CONTROLOPTS ||
322 last->inp_socket->so_options & SO_TIMESTAMP)
323 ip6_savecontrol(last, m, &opts);
324 /* Strip intermediate headers. */
325 m_adj(m, *offp);
326 if (sbappendaddr(&last->inp_socket->so_rcv,
327 (struct sockaddr *)&fromsa, m, opts) == 0) {
328 soroverflow(last->inp_socket);
329 m_freem(m);
330 if (opts)
331 m_freem(opts);
332 RIP6STAT_INC(rip6s_fullsock);
333 } else
334 sorwakeup(last->inp_socket);
335 INP_RUNLOCK(last);
336 } else {
337 RIP6STAT_INC(rip6s_nosock);
338 if (m->m_flags & M_MCAST)
339 RIP6STAT_INC(rip6s_nosockmcast);
340 if (proto == IPPROTO_NONE)
341 m_freem(m);
342 else
343 icmp6_error(m, ICMP6_PARAM_PROB,
344 ICMP6_PARAMPROB_NEXTHEADER,
345 ip6_get_prevhdr(m, *offp));
346 IP6STAT_DEC(ip6s_delivered);
347 }
348 return (IPPROTO_DONE);
349 }
350
351 void
rip6_ctlinput(int cmd,struct sockaddr * sa,void * d)352 rip6_ctlinput(int cmd, struct sockaddr *sa, void *d)
353 {
354 struct ip6ctlparam *ip6cp = NULL;
355 const struct sockaddr_in6 *sa6_src = NULL;
356 void *cmdarg;
357 struct inpcb *(*notify)(struct inpcb *, int) = in6_rtchange;
358
359 if (sa->sa_family != AF_INET6 ||
360 sa->sa_len != sizeof(struct sockaddr_in6))
361 return;
362
363 if ((unsigned)cmd >= PRC_NCMDS)
364 return;
365 if (PRC_IS_REDIRECT(cmd))
366 notify = in6_rtchange, d = NULL;
367 else if (cmd == PRC_HOSTDEAD)
368 d = NULL;
369 else if (inet6ctlerrmap[cmd] == 0)
370 return;
371
372 /*
373 * If the parameter is from icmp6, decode it.
374 */
375 if (d != NULL) {
376 ip6cp = (struct ip6ctlparam *)d;
377 cmdarg = ip6cp->ip6c_cmdarg;
378 sa6_src = ip6cp->ip6c_src;
379 } else {
380 cmdarg = NULL;
381 sa6_src = &sa6_any;
382 }
383
384 (void) in6_pcbnotify(&V_ripcbinfo, sa, 0,
385 (const struct sockaddr *)sa6_src, 0, cmd, cmdarg, notify);
386 }
387
388 /*
389 * Generate IPv6 header and pass packet to ip6_output. Tack on options user
390 * may have setup with control call.
391 */
392 int
rip6_output(struct mbuf * m,struct socket * so,...)393 rip6_output(struct mbuf *m, struct socket *so, ...)
394 {
395 struct mbuf *control;
396 struct m_tag *mtag;
397 struct sockaddr_in6 *dstsock;
398 struct ip6_hdr *ip6;
399 struct inpcb *inp;
400 u_int plen = m->m_pkthdr.len;
401 int error = 0;
402 struct ip6_pktopts opt, *optp;
403 struct ifnet *oifp = NULL;
404 int type = 0, code = 0; /* for ICMPv6 output statistics only */
405 int scope_ambiguous = 0;
406 int use_defzone = 0;
407 int hlim = 0;
408 struct in6_addr in6a;
409 va_list ap;
410
411 va_start(ap, so);
412 dstsock = va_arg(ap, struct sockaddr_in6 *);
413 control = va_arg(ap, struct mbuf *);
414 va_end(ap);
415
416 inp = sotoinpcb(so);
417 INP_WLOCK(inp);
418
419 if (control != NULL) {
420 if ((error = ip6_setpktopts(control, &opt,
421 inp->in6p_outputopts, so->so_cred,
422 so->so_proto->pr_protocol)) != 0) {
423 goto bad;
424 }
425 optp = &opt;
426 } else
427 optp = inp->in6p_outputopts;
428
429 /*
430 * Check and convert scope zone ID into internal form.
431 *
432 * XXX: we may still need to determine the zone later.
433 */
434 if (!(so->so_state & SS_ISCONNECTED)) {
435 if (!optp || !optp->ip6po_pktinfo ||
436 !optp->ip6po_pktinfo->ipi6_ifindex)
437 use_defzone = V_ip6_use_defzone;
438 if (dstsock->sin6_scope_id == 0 && !use_defzone)
439 scope_ambiguous = 1;
440 if ((error = sa6_embedscope(dstsock, use_defzone)) != 0)
441 goto bad;
442 }
443
444 /*
445 * For an ICMPv6 packet, we should know its type and code to update
446 * statistics.
447 */
448 if (so->so_proto->pr_protocol == IPPROTO_ICMPV6) {
449 struct icmp6_hdr *icmp6;
450 if (m->m_len < sizeof(struct icmp6_hdr) &&
451 (m = m_pullup(m, sizeof(struct icmp6_hdr))) == NULL) {
452 error = ENOBUFS;
453 goto bad;
454 }
455 icmp6 = mtod(m, struct icmp6_hdr *);
456 type = icmp6->icmp6_type;
457 code = icmp6->icmp6_code;
458 }
459
460 M_PREPEND(m, sizeof(*ip6), M_NOWAIT);
461 if (m == NULL) {
462 error = ENOBUFS;
463 goto bad;
464 }
465 ip6 = mtod(m, struct ip6_hdr *);
466
467 /*
468 * Source address selection.
469 */
470 error = in6_selectsrc_socket(dstsock, optp, inp, so->so_cred,
471 scope_ambiguous, &in6a, &hlim);
472
473 if (error)
474 goto bad;
475 error = prison_check_ip6(inp->inp_cred, &in6a);
476 if (error != 0)
477 goto bad;
478 ip6->ip6_src = in6a;
479
480 ip6->ip6_dst = dstsock->sin6_addr;
481
482 /*
483 * Fill in the rest of the IPv6 header fields.
484 */
485 ip6->ip6_flow = (ip6->ip6_flow & ~IPV6_FLOWINFO_MASK) |
486 (inp->inp_flow & IPV6_FLOWINFO_MASK);
487 ip6->ip6_vfc = (ip6->ip6_vfc & ~IPV6_VERSION_MASK) |
488 (IPV6_VERSION & IPV6_VERSION_MASK);
489
490 /*
491 * ip6_plen will be filled in ip6_output, so not fill it here.
492 */
493 ip6->ip6_nxt = inp->inp_ip_p;
494 ip6->ip6_hlim = hlim;
495
496 if (so->so_proto->pr_protocol == IPPROTO_ICMPV6 ||
497 inp->in6p_cksum != -1) {
498 struct mbuf *n;
499 int off;
500 u_int16_t *p;
501
502 /* Compute checksum. */
503 if (so->so_proto->pr_protocol == IPPROTO_ICMPV6)
504 off = offsetof(struct icmp6_hdr, icmp6_cksum);
505 else
506 off = inp->in6p_cksum;
507 if (plen < off + 2) {
508 error = EINVAL;
509 goto bad;
510 }
511 off += sizeof(struct ip6_hdr);
512
513 n = m;
514 while (n && n->m_len <= off) {
515 off -= n->m_len;
516 n = n->m_next;
517 }
518 if (!n)
519 goto bad;
520 p = (u_int16_t *)(mtod(n, caddr_t) + off);
521 *p = 0;
522 *p = in6_cksum(m, ip6->ip6_nxt, sizeof(*ip6), plen);
523 }
524
525 /*
526 * Send RA/RS messages to user land for protection, before sending
527 * them to rtadvd/rtsol.
528 */
529 if ((send_sendso_input_hook != NULL) &&
530 so->so_proto->pr_protocol == IPPROTO_ICMPV6) {
531 switch (type) {
532 case ND_ROUTER_ADVERT:
533 case ND_ROUTER_SOLICIT:
534 mtag = m_tag_get(PACKET_TAG_ND_OUTGOING,
535 sizeof(unsigned short), M_NOWAIT);
536 if (mtag == NULL)
537 goto bad;
538 m_tag_prepend(m, mtag);
539 }
540 }
541
542 error = ip6_output(m, optp, NULL, 0, inp->in6p_moptions, &oifp, inp);
543 if (so->so_proto->pr_protocol == IPPROTO_ICMPV6) {
544 if (oifp)
545 icmp6_ifoutstat_inc(oifp, type, code);
546 ICMP6STAT_INC(icp6s_outhist[type]);
547 } else
548 RIP6STAT_INC(rip6s_opackets);
549
550 goto freectl;
551
552 bad:
553 if (m)
554 m_freem(m);
555
556 freectl:
557 if (control != NULL) {
558 ip6_clearpktopts(&opt, -1);
559 m_freem(control);
560 }
561 INP_WUNLOCK(inp);
562 return (error);
563 }
564
565 /*
566 * Raw IPv6 socket option processing.
567 */
568 int
rip6_ctloutput(struct socket * so,struct sockopt * sopt)569 rip6_ctloutput(struct socket *so, struct sockopt *sopt)
570 {
571 struct inpcb *inp;
572 int error;
573
574 if (sopt->sopt_level == IPPROTO_ICMPV6)
575 /*
576 * XXX: is it better to call icmp6_ctloutput() directly
577 * from protosw?
578 */
579 return (icmp6_ctloutput(so, sopt));
580 else if (sopt->sopt_level != IPPROTO_IPV6) {
581 if (sopt->sopt_level == SOL_SOCKET &&
582 sopt->sopt_name == SO_SETFIB) {
583 inp = sotoinpcb(so);
584 INP_WLOCK(inp);
585 inp->inp_inc.inc_fibnum = so->so_fibnum;
586 INP_WUNLOCK(inp);
587 return (0);
588 }
589 return (EINVAL);
590 }
591
592 error = 0;
593
594 switch (sopt->sopt_dir) {
595 case SOPT_GET:
596 switch (sopt->sopt_name) {
597 case MRT6_INIT:
598 case MRT6_DONE:
599 case MRT6_ADD_MIF:
600 case MRT6_DEL_MIF:
601 case MRT6_ADD_MFC:
602 case MRT6_DEL_MFC:
603 case MRT6_PIM:
604 error = ip6_mrouter_get ? ip6_mrouter_get(so, sopt) :
605 EOPNOTSUPP;
606 break;
607 case IPV6_CHECKSUM:
608 error = ip6_raw_ctloutput(so, sopt);
609 break;
610 default:
611 error = ip6_ctloutput(so, sopt);
612 break;
613 }
614 break;
615
616 case SOPT_SET:
617 switch (sopt->sopt_name) {
618 case MRT6_INIT:
619 case MRT6_DONE:
620 case MRT6_ADD_MIF:
621 case MRT6_DEL_MIF:
622 case MRT6_ADD_MFC:
623 case MRT6_DEL_MFC:
624 case MRT6_PIM:
625 error = ip6_mrouter_set ? ip6_mrouter_set(so, sopt) :
626 EOPNOTSUPP;
627 break;
628 case IPV6_CHECKSUM:
629 error = ip6_raw_ctloutput(so, sopt);
630 break;
631 default:
632 error = ip6_ctloutput(so, sopt);
633 break;
634 }
635 break;
636 }
637
638 return (error);
639 }
640
641 static int
rip6_attach(struct socket * so,int proto,struct thread * td)642 rip6_attach(struct socket *so, int proto, struct thread *td)
643 {
644 struct inpcb *inp;
645 struct icmp6_filter *filter;
646 int error;
647
648 inp = sotoinpcb(so);
649 KASSERT(inp == NULL, ("rip6_attach: inp != NULL"));
650
651 error = priv_check(td, PRIV_NETINET_RAW);
652 if (error)
653 return (error);
654 error = soreserve(so, rip_sendspace, rip_recvspace);
655 if (error)
656 return (error);
657 filter = malloc(sizeof(struct icmp6_filter), M_PCB, M_NOWAIT);
658 if (filter == NULL)
659 return (ENOMEM);
660 INP_INFO_WLOCK(&V_ripcbinfo);
661 error = in_pcballoc(so, &V_ripcbinfo);
662 if (error) {
663 INP_INFO_WUNLOCK(&V_ripcbinfo);
664 free(filter, M_PCB);
665 return (error);
666 }
667 inp = (struct inpcb *)so->so_pcb;
668 INP_INFO_WUNLOCK(&V_ripcbinfo);
669 inp->inp_vflag |= INP_IPV6;
670 inp->inp_ip_p = (long)proto;
671 inp->in6p_hops = -1; /* use kernel default */
672 inp->in6p_cksum = -1;
673 inp->in6p_icmp6filt = filter;
674 ICMP6_FILTER_SETPASSALL(inp->in6p_icmp6filt);
675 INP_WUNLOCK(inp);
676 return (0);
677 }
678
679 static void
rip6_detach(struct socket * so)680 rip6_detach(struct socket *so)
681 {
682 struct inpcb *inp;
683
684 inp = sotoinpcb(so);
685 KASSERT(inp != NULL, ("rip6_detach: inp == NULL"));
686
687 if (so == V_ip6_mrouter && ip6_mrouter_done)
688 ip6_mrouter_done();
689 /* xxx: RSVP */
690 INP_INFO_WLOCK(&V_ripcbinfo);
691 INP_WLOCK(inp);
692 free(inp->in6p_icmp6filt, M_PCB);
693 in_pcbdetach(inp);
694 in_pcbfree(inp);
695 INP_INFO_WUNLOCK(&V_ripcbinfo);
696 }
697
698 /* XXXRW: This can't ever be called. */
699 static void
rip6_abort(struct socket * so)700 rip6_abort(struct socket *so)
701 {
702 struct inpcb *inp;
703
704 inp = sotoinpcb(so);
705 KASSERT(inp != NULL, ("rip6_abort: inp == NULL"));
706
707 soisdisconnected(so);
708 }
709
710 static void
rip6_close(struct socket * so)711 rip6_close(struct socket *so)
712 {
713 struct inpcb *inp;
714
715 inp = sotoinpcb(so);
716 KASSERT(inp != NULL, ("rip6_close: inp == NULL"));
717
718 soisdisconnected(so);
719 }
720
721 static int
rip6_disconnect(struct socket * so)722 rip6_disconnect(struct socket *so)
723 {
724 struct inpcb *inp;
725
726 inp = sotoinpcb(so);
727 KASSERT(inp != NULL, ("rip6_disconnect: inp == NULL"));
728
729 if ((so->so_state & SS_ISCONNECTED) == 0)
730 return (ENOTCONN);
731 inp->in6p_faddr = in6addr_any;
732 rip6_abort(so);
733 return (0);
734 }
735
736 static int
rip6_bind(struct socket * so,struct sockaddr * nam,struct thread * td)737 rip6_bind(struct socket *so, struct sockaddr *nam, struct thread *td)
738 {
739 struct inpcb *inp;
740 struct sockaddr_in6 *addr = (struct sockaddr_in6 *)nam;
741 struct ifaddr *ifa = NULL;
742 int error = 0;
743
744 inp = sotoinpcb(so);
745 KASSERT(inp != NULL, ("rip6_bind: inp == NULL"));
746
747 if (nam->sa_len != sizeof(*addr))
748 return (EINVAL);
749 if ((error = prison_check_ip6(td->td_ucred, &addr->sin6_addr)) != 0)
750 return (error);
751 if (CK_STAILQ_EMPTY(&V_ifnet) || addr->sin6_family != AF_INET6)
752 return (EADDRNOTAVAIL);
753 if ((error = sa6_embedscope(addr, V_ip6_use_defzone)) != 0)
754 return (error);
755
756 NET_EPOCH_ENTER();
757 if (!IN6_IS_ADDR_UNSPECIFIED(&addr->sin6_addr) &&
758 (ifa = ifa_ifwithaddr((struct sockaddr *)addr)) == NULL) {
759 NET_EPOCH_EXIT();
760 return (EADDRNOTAVAIL);
761 }
762 if (ifa != NULL &&
763 ((struct in6_ifaddr *)ifa)->ia6_flags &
764 (IN6_IFF_ANYCAST|IN6_IFF_NOTREADY|
765 IN6_IFF_DETACHED|IN6_IFF_DEPRECATED)) {
766 NET_EPOCH_EXIT();
767 return (EADDRNOTAVAIL);
768 }
769 NET_EPOCH_EXIT();
770 INP_INFO_WLOCK(&V_ripcbinfo);
771 INP_WLOCK(inp);
772 inp->in6p_laddr = addr->sin6_addr;
773 INP_WUNLOCK(inp);
774 INP_INFO_WUNLOCK(&V_ripcbinfo);
775 return (0);
776 }
777
778 static int
rip6_connect(struct socket * so,struct sockaddr * nam,struct thread * td)779 rip6_connect(struct socket *so, struct sockaddr *nam, struct thread *td)
780 {
781 struct inpcb *inp;
782 struct sockaddr_in6 *addr = (struct sockaddr_in6 *)nam;
783 struct in6_addr in6a;
784 int error = 0, scope_ambiguous = 0;
785
786 inp = sotoinpcb(so);
787 KASSERT(inp != NULL, ("rip6_connect: inp == NULL"));
788
789 if (nam->sa_len != sizeof(*addr))
790 return (EINVAL);
791 if (CK_STAILQ_EMPTY(&V_ifnet))
792 return (EADDRNOTAVAIL);
793 if (addr->sin6_family != AF_INET6)
794 return (EAFNOSUPPORT);
795
796 /*
797 * Application should provide a proper zone ID or the use of default
798 * zone IDs should be enabled. Unfortunately, some applications do
799 * not behave as it should, so we need a workaround. Even if an
800 * appropriate ID is not determined, we'll see if we can determine
801 * the outgoing interface. If we can, determine the zone ID based on
802 * the interface below.
803 */
804 if (addr->sin6_scope_id == 0 && !V_ip6_use_defzone)
805 scope_ambiguous = 1;
806 if ((error = sa6_embedscope(addr, V_ip6_use_defzone)) != 0)
807 return (error);
808
809 INP_INFO_WLOCK(&V_ripcbinfo);
810 INP_WLOCK(inp);
811 /* Source address selection. XXX: need pcblookup? */
812 error = in6_selectsrc_socket(addr, inp->in6p_outputopts,
813 inp, so->so_cred, scope_ambiguous, &in6a, NULL);
814 if (error) {
815 INP_WUNLOCK(inp);
816 INP_INFO_WUNLOCK(&V_ripcbinfo);
817 return (error);
818 }
819
820 inp->in6p_faddr = addr->sin6_addr;
821 inp->in6p_laddr = in6a;
822 soisconnected(so);
823 INP_WUNLOCK(inp);
824 INP_INFO_WUNLOCK(&V_ripcbinfo);
825 return (0);
826 }
827
828 static int
rip6_shutdown(struct socket * so)829 rip6_shutdown(struct socket *so)
830 {
831 struct inpcb *inp;
832
833 inp = sotoinpcb(so);
834 KASSERT(inp != NULL, ("rip6_shutdown: inp == NULL"));
835
836 INP_WLOCK(inp);
837 socantsendmore(so);
838 INP_WUNLOCK(inp);
839 return (0);
840 }
841
842 static int
rip6_send(struct socket * so,int flags,struct mbuf * m,struct sockaddr * nam,struct mbuf * control,struct thread * td)843 rip6_send(struct socket *so, int flags, struct mbuf *m, struct sockaddr *nam,
844 struct mbuf *control, struct thread *td)
845 {
846 struct inpcb *inp;
847 struct sockaddr_in6 tmp;
848 struct sockaddr_in6 *dst;
849 int ret;
850
851 inp = sotoinpcb(so);
852 KASSERT(inp != NULL, ("rip6_send: inp == NULL"));
853
854 /* Always copy sockaddr to avoid overwrites. */
855 /* Unlocked read. */
856 if (so->so_state & SS_ISCONNECTED) {
857 if (nam) {
858 m_freem(m);
859 return (EISCONN);
860 }
861 /* XXX */
862 bzero(&tmp, sizeof(tmp));
863 tmp.sin6_family = AF_INET6;
864 tmp.sin6_len = sizeof(struct sockaddr_in6);
865 INP_RLOCK(inp);
866 bcopy(&inp->in6p_faddr, &tmp.sin6_addr,
867 sizeof(struct in6_addr));
868 INP_RUNLOCK(inp);
869 dst = &tmp;
870 } else {
871 if (nam == NULL) {
872 m_freem(m);
873 return (ENOTCONN);
874 }
875 if (nam->sa_len != sizeof(struct sockaddr_in6)) {
876 m_freem(m);
877 return (EINVAL);
878 }
879 tmp = *(struct sockaddr_in6 *)nam;
880 dst = &tmp;
881
882 if (dst->sin6_family == AF_UNSPEC) {
883 /*
884 * XXX: we allow this case for backward
885 * compatibility to buggy applications that
886 * rely on old (and wrong) kernel behavior.
887 */
888 log(LOG_INFO, "rip6 SEND: address family is "
889 "unspec. Assume AF_INET6\n");
890 dst->sin6_family = AF_INET6;
891 } else if (dst->sin6_family != AF_INET6) {
892 m_freem(m);
893 return(EAFNOSUPPORT);
894 }
895 }
896 ret = rip6_output(m, so, dst, control);
897 return (ret);
898 }
899
900 struct pr_usrreqs rip6_usrreqs = {
901 .pru_abort = rip6_abort,
902 .pru_attach = rip6_attach,
903 .pru_bind = rip6_bind,
904 .pru_connect = rip6_connect,
905 .pru_control = in6_control,
906 .pru_detach = rip6_detach,
907 .pru_disconnect = rip6_disconnect,
908 .pru_peeraddr = in6_getpeeraddr,
909 .pru_send = rip6_send,
910 .pru_shutdown = rip6_shutdown,
911 .pru_sockaddr = in6_getsockaddr,
912 .pru_close = rip6_close,
913 };
914